SEPARATION FROM LOVED ONES IN THEFEAR OF DEATH

DEBRA M. BATH

School of Psychology, Griffith University, Gold Coast,Queensland, Australia

Individuals’ death anxiety or fear of death has been extensively investigated, andthere are numerous conceptualizations used in the literature, including a distinc-tion between the dimensions of death and dying of self, and death and dying ofothers. This article addresses a gap in the literature and re-examines the relation-ship between these two dimensions, which are assumed to be positively, linearlyrelated. Using both quantitative and qualitative data, this study indicates thatregardless of the degree to which individuals fear their own death, most indivi-duals fear the death and dying of others. Specifically, the leaving, or loss of lovedones, was a central theme in people’s fear of death, and this is discussed inrelation to current trends in the literature.

In recent decades research on death attitudes, including fear ofdeath, death anxiety, and death competency, has flourished. Muchof this research has been directed at establishing measures of suchconstructs, and refining these measures in terms of reliability,validity, and soundness of factor structures (Neimeyer, Moser, &Wittkowski, 2003). The use of such instruments in different clinicaland occupational settings has also been prevalent. There are now avariety of instruments used to measure death fear and=or deathanxiety, and whilst early work tended to be unidimensional (e.g.,Handal, 1975; McMordie, 1979; Templer, 1970) more recentlydeveloped instruments recognize the multidimensional nature ofsuch constructs (Collett & Lester, 1969; Florian & Kravetz, 1983;Hoelter, 1979; Leming & Dickinson, 1985; Wittkowski, 2001).For example, the Multidimensional Fear of Death Scale (MFODS;

Received 22 September 2007; accepted 1 December 2008.Address correspondence to Debra M. Bath, PhD, School of Psychology, Griffith

University, Gold Coast PMB 50, Gold Coast Mail Centre Gold Coast, Queensland 9726,Australia. E-mail: d.bath@griffith.edu.au

Death Studies, 34: 404–425, 2010Copyright # Taylor & Francis Group, LLCISSN: 0748-1187 print=1091-7683 onlineDOI: 10.1080/07481181003697639

404

Hoelter, 1979) includes eight factors: Fear of the Dying Process,Fear of the Dead, Fear of Being Destroyed, Fear for SignificantOthers, Fear of the Unknown, Fear of Conscious Death, Fear forthe Body After Death, and Fear of Premature Death. The Fear ofPersonal Death Scale (FPDS; Florian & Kravetz, 1983) further dis-tinguishes between intrapersonal concerns (loss of self-fulfilment,self-annihilation), interpersonal concerns (loss of social identity,consequences to family and friends), and transpersonal concerns(transcendental consequences, punishment in the hereafter).

However, the Collett–Lester Fear of Death Scale (CLFD;Collett & Lester, 1969) is perhaps the most commonly used instru-ment that clearly and systematically distinguishes between two keydimensions involving death: (a) the state of death versus the pro-cess of dying and (b) one’s own death versus the death of others.As such it includes four subscales: Death of Self (e.g., total isolationof death, shortness of life, never thinking or experiencing), Dyingof Self (e.g., pain involved in dying, intellectual degeneration, lackof control over process, grief of others), Death of Others (e.g., los-ing someone close, never being able to communicate again, feelinglonely without the person), and Dying of Others (e.g., watching theperson suffer, having to be with someone who is dying). Since itsoriginal development, the CLFD has been revised, reducing thenumber of items from 36 to 28 in order to have equal weightingacross subscales for scoring purposes and to remove problematicor deviant items (Lester, 1994; Lester & Abdel-Khalek, 2003).However, there are differing views regarding the factorial validityof this scale and therefore about the usage of the scale in its pro-posed four-factor structure. Typically, factor analysis of the scaleitems reveals five to seven factors with cross-loadings for severalitems, particularly for the Death of Others and Dying of Otherssubscales (Mooney & O’Gorman, 2001; Neimeyer et al., 2003),which tends to indicate factorial instability and thus equivocal sup-port for the validity of the proposed constructs. It has been pro-posed, therefore, that researchers should consider disregarding its‘‘attractive and symmetrical factor structure’’ (Neimeyer et al.,2003, p. 55) and instead use a factor structure empirically derivedeven though it may be more difficult to interpret. Conversely, ithas also been argued that because of the natural associationbetween fear of own death and fear of others, and fear of deathand fear of dying, such a factor pattern is likely and it does not

Fear of Death and Loss of Loved Ones 405

contradict the existing allocation of items to each of the foursubscales on a meaningful basis (Lester, 2004). Despite a poten-tially problematic factor structure, it has been used extensively ina wide range of research areas, and more recently has beenadapted for use with languages other than English (Abdel-Khalek& Lester, 2004; Tomas-Sabado, Limonero, & Abdel-Khalek,2007; Yasukawa, 2006).

Research using the CLFD has examined a wide range of vari-ables in association with fear of death such as age, gender, religi-osity, depression, general anxiety, extraversion, neuroticism, andattitudes toward suicide (for an overview, see Lester, 1994). Studieshave often shown mixed findings; for example, while some havereported that CLFD scores are lower for older adults compared withyoung adults (Smith, Nehemkis & Charter, 1983–1984) and womenreport higher fear of death than men (Neimeyer, Bagley, & Moore,1986; Robbins, 1990), other studies have found little differencebetween these groups (e.g., Lester, 1972, 1984–1985; Loo, 1984).

The CLFD has also been used in evaluating the impact oftraining programs on participants’ fear of death, such as palliativecare volunteer training (Claxton-Oldfield, Crain, & Claxton-Oldfield, 2007; Hayslip & Walling, 1985–1986) and death edu-cation for nursing students (Mooney, 2005; Mooney & O’Gorman,2001). It has also been used to examine levels of death anxiety in avariety of populations such as widows (Straub & Roberts, 2001),the elderly (Clements, 1998), and organ donors (Robbins, 1990),as well as with practitioner groups involved with death and dyingsuch as medical students (Jordan, Ellis, & Grallo, 1986), psycholo-gists (Soghomonian, 2000), suicide intervention workers(Neimeyer & Dingemans, 1980–1981), and trainee counsellors(Larrabee, 1978).

Psychometric studies of other death anxiety instruments andrelated variables such as death competency (Robbins, 1990–1991) have also used the CLFD to establish validity (e.g.,Abdel-Khalek, 2002a). For example, Robbins found that theCoping with Death Scale was negatively associated with theCLFD, which provided some degree of convergent validity forthe measure. Vargo (1980) found that correlations between theTempler Death Anxiety Scale (DAS; Templer, 1970) and theCLFD were highest for the fear of own death and dying subscales,which indicated concurrent validity for the DAS.

406 D. M. Bath

Examination of the relationships amongst the four CLFDsubscales has tended to show moderately strong positive correla-tions between the different dimensions. For example, Mooneyand O’Gorman (2001) reported correlations of .54 between Deathof Self and Dying of Self, .49 between Death of Self and Death ofOthers, .51 between Death of Self and Dying of Others, .49between Dying of Self and Death of Others, .61 between Dyingof Self and Dying of Others, and .67 between Death of Othersand Dying of Others. These results are similar to more recentanalyses by Lester (2004), although the correlations tend to behigher than some previous studies (see Lester, 1994). Such relation-ships indicate that, for example, individuals who score high on thefear of Death of Self will tend also to score high on the fear ofDeath of Others, and vice versa, those who score low on fear ofDeath of Self will also score low on fear of Death of Others.However, given the magnitude of reported correlations, there areobviously many individuals who do not fit this relationship.

It is the aim of the present study to examine this issue in orderto provide a richer explanation of this relationship between the fearof one’s own death and dying and that of others. The approachtaken combines both quantitative and qualitative data and analyses.First, the pattern of scores across the CLFD scales for those showinghigh fear of own death compared with those showing low fear ofown death will be examined to provide a richer profile of these rela-tionships. This will go beyond establishing a basic positive relation-ship as done in previous correlational analyses. Second, usingqualitative responses regarding what aspects of death are fearedmost, it is hoped to provide additional clarification and expositionof the observed patterns across CLFD scales. Given the exploratorynature of the study, no hypotheses were made a priori.

Method

Participants

Ninety-one students enrolled in an introductory psychology unitvolunteered to participate in this study. The average age was20.8 (SD¼ 6.41) ranging from 16 to 54 years, and there were 23(25%) men and 68 (75%) women. Most participants wereAustralian (89%) with the remaining participants identifying as

Fear of Death and Loss of Loved Ones 407

North American, English, Swedish, South African, Filipino, orTaiwanese. Thirty-six percent of participants were religious, 41%were not religious, and 23%were unsure of their religiosity. Of thosewho identified as being religious, 86% reported being of theChristian faith (21% of these specified being Catholic), with theremaining few participants either not specifying their faith or speci-fying a non-mainstream faith (e.g., Jehovah’s Witness, New Age).

Materials and Procedure

A questionnaire that included both qualitative and quantitativedata was developed for the study, and included general demo-graphics (age, gender, nationality, and religion), questions aboutpast experiences of death, followed by a section on fear of death.

After completing the demographic questions, participantswere asked if they had ever lost a loved one and to indicate theirrelationship to this person, how old they were when the persondied, and how close they were to this person on a scale from 1(not at all close) to 5 (extremely close). (Other questions relating towhether participants had ever known someone who was grievingwere also asked but were not included in this study).

Participants were then asked to respond to a generalopen-ended question regarding their fear of death: ‘‘Many peoplefear death. From your point of view, which aspects of death do youfind the most worrying or fearsome?’’ Following this question, therevised version of the CLFD was presented. The CLFD scaleincludes 4 subscales, each comprising 7 items: Death of Self, Dyingof Self, Death of Others, and Dying of Others. Respondents areasked to consider ‘‘How disturbed or made anxious are you bythe following aspects of death and dying?’’ and to rate each itemusing a 5-point Likert scale from 1 (not) 5 (very). The revised versionhas shown improved factor-structure, with stronger support for thefour proposed subscales than the original instrument, and goodtest–retest reliability with Cronbach’s alpha ranging between .88and .92 (Lester & Abdel-Khalek, 2003; Mooney & O’Gorman,2001). Despite the arguments made regarding the factorial insta-bility of the CLFD, the four subscales were retained for the presentstudy in order to provide greater exploratory richness in examin-ing participants’ fear of death and for comparability with previousresearch.

408 D. M. Bath

The questionnaire was completed online via a secure siteadministered by the university in which the participants wereenrolled. Potential participants were invited to participate in thestudy via an information sheet displayed as part of a studentresearch participation program. Participants were provided witha web address for the questionnaire and instructions for gainingsecure access with anonymity. The questionnaire took appropri-ately 20 to 30min to complete.

Results

Participant Profile

Most participants had experienced the loss of a loved one (82%) andthe closeness of their relationship to the deceased was generally‘‘moderately close’’ (M¼ 3.53, SD¼ 1.17). The deceased personwas primarily a grandparent (40%), friend (20%) or extended familymember (19%), although for some participants (16%) this was animmediate family member (parent, sibling, or partner). The age atwhich participants experienced the death of a loved one ranged from5 to 45, with the majority of people being in their late teens (48%between 15 and 17 years). These characteristics are typical for thisage group and community context (e.g., Knight, Elfenbein, &Capozzi, 2000; Vickio, Cavanaugh, & Attig, 1990;Woodford, 2002).

CLFD

Mean scores, bivariate correlations, and alpha coefficients (in par-entheses) for the four subscales of the CLFD are presented inTable 1. Scale reliabilities and intercorrelations are consistent inmagnitude with previous studies (e.g., Lester & Abdel-Khalek,2003; Mooney & O’Gorman, 2001). Previous analyses have usedadditive scale score calculations and so for comparability totalscores (possible range 7 to 35) using this method are reported here.However, the scale mean scores (possible range 1 to 5) are alsoreported for ease of interpretation. There were no significant corre-lations between age and CLFD scales, although it should be notedthat the sample did not include participants older than 54 and fewwere older than 35 years.

Fear of Death and Loss of Loved Ones 409

Gender differences were examined using a 2� 4 mixed analy-sis of variance (ANOVA) with gender as a between-subjects factorand CLFD as a within-subjects factor. A main effect of CLFD wasevident, F (3, 243)¼ 27.35, p< .0001 (gp

2¼ .25) with scores onDeath of Others being significantly higher than all other subscales(p< .0001) and no significant differences between fear of Death ofSelf, Dying of Self, and Dying of Others. There was also a maineffect of gender, F (1, 81)¼ 31.19, p< .0001 (gp

2 ¼ .28), withwomen showing greater fear in general than men. These effectswere qualified by a significant gender by CLFD interaction, F (3,243)¼ 4.96, p< .01 (gp

2 ¼ .06). For men, fear of Death of Self,Dying of Self and Dying of Others were not significantly differentfrom one another, but were all significantly lower than fear ofDeath of Others (all p< .001). However for women, fear of Deathof Self was significantly lower than Dying of Self and Dying ofOthers and these three subscales were significantly lower than fearof Death of Others (all p< .01). As can be seen in Figure 1, thedistinction between men and women is marked on all subscalesexcept Death of Others where both men and women are mostfearful and are most similar to one another.

To further examine fear of death in terms of self and others,participant scores on the four subscales were recoded into 2 cate-gories, high fear (above the mean) and low fear (below the mean).Given that correlations (Table 1) show moderate to strong positiverelationships between the Death=Dying of Self scales and the

TABLE 1 Descriptive Statistics, Alpha Coefficients, and Intercorrelations for theCLFD Subscales

Subscale 1 2 3 4

1. Own death (.83)2. Own dying .65 (.89)3. Others death .56 .53 (.71)4. Others dying .44 .63 .56 (.87)

M total score 21.33 (6.58) 22.71 (6.80) 26.94 (4.81) 24.01 (6.03)Total for men 18.19 (7.67) 17.00 (5.67) 24.62 (4.71) 17.95 (6.14)Total for women 22.33 (5.91) 24.56 (6.14) 27.68 (4.64) 25.77 (4.74)

M score 3.05 (0.94) 3.24 (0.98) 3.85 (0.69) 3.43 (0.86)M for men 2.60 (1.09) 2.42 (0.81) 3.52 (0.67) 2.56 (0.88)M for women 3.19 (0.85) 3.51 (0.88) 3.95 (0.66) 3.68 (0.68)

410 D. M. Bath

Death=Dying of Others scales, it can be assumed that participantsfalling into the high fear category of one dimension (i.e., own)would also fall into the high fear category of the other dimension(i.e., others), and vice versa in terms of low fear.

This assumption was tested using a series of 2 (high, low)� 2(high, low) Pearson’s chi-square analyses, with Bonferonni adjust-ment for multiple tests to control for type 1 error rate (a¼ .01):(a) Death of Self�Dying of Self; Death of Self�Death of Others;Death of Self�Dying of Others, and (b) Dying of Self�Death ofOthers; Dying of Self�Dying of Others. Results are presented inFigures 2 and 3. For each analysis, if the assumed pattern of

FIGURE 1 Interaction between gender and CLFD scales.

FIGURE 2 Interaction between fear of own death and other CLFD scales.

Fear of Death and Loss of Loved Ones 411

relationships exists then the results should be significant, indicatingthe there is an unequal proportion of individuals falling into thecross-tabulations.

As expected, analyses revealed that the interaction betweenDeath of Self�Dying of Self was significant, v2(1)¼ 30.47,p< .0001, showing that those high in fear of own death werealso most likely to be high in fear of own dying (87.2%) and thoselow in fear of own death were also most likely to be low in fear ofown dying (72.3%). A similar pattern, though much less marked,was observed for the interaction between Death of Self�Deathof Others, v2(1)¼ 19.69, p< .001, with those high in fear of owndeath also most likely to be high in fear of others death (84.6%)but only a small majority of those low in fear of own death alsobeing low in fear of others death (60%). This pattern was even lessevident, and not significant, for the Death of Self�Dying ofOthers analyses, v2(1)¼ 6.42, p> .01, indicating a more equaldistribution across categories. As can be seen in Figure 2, thosehigh in fear of own death still tended to be high in fear of othersdying, but those low in fear of own death were much more equallydistributed across high and low fear of others dying (43.5% and56.5%).

Analysis examining the relationship between Dying of Self�Death of Others was significant, v2(1)¼ 16.19, p¼ .0001. Similarly,the interaction between Dying of Self�Dying of Others was sig-nificant, v2(1)¼ 15.91, p¼ .0001. As can be seen in Figure 3, part-icipants who were high in fear of own dying were also generallyhigh in fear of others dying and others death. Again, this patternwas evident but much less marked for those low in fear of owndying, where there was a greater tendency for some of these part-icipants to be in the high fear (rather than low fear) of others deathand dying categories.

Together these analyses indicate that, in general, personswith a high fear of their own death (or own dying) are mostlikely to show a high rather than low fear of others death anddying. On the other hand, persons with a low fear of theirown death (or own dying) do not show the same clear relation-ship with fear of others death and dying. Instead of the majorityof these persons showing a low fear of others death and dying,they are more equally dispersed as high or low in fear of othersdeath and dying.

412 D. M. Bath

Qualitative Data

Responses to the question ‘‘. . .which aspects of death do you findthe most worrying or fearsome?’’ may reveal some insights into thisrelationship between fear of own death and fear of others death.These data were analyzed using thematic analysis, a method of ana-lyzing qualitative data which involves sorting the information, iden-tifying recurrent ideas or topics that occur in the data and thusestablishing themes. The same theme may be manifest in differentways, and may include sub-themes, so it is vital that the process oftheme identification and clarification involves re-examination ofthe data after themes have been identified in order to search foradditional occurrences or representations, and verify the themes(Hayes, 2002). First, I analyzed the entire set of responses for recur-rent themes. A colleague then examined the responses for evidenceof any additional themes or problematic themes. Once a set ofthemes was agreed upon, both independently coded the set ofresponses. Inter-rater reliability was 92%. Responses that initiallydiffered in coding were discussed and a final code agreed upon.

Analysis of responses revealed themes that can be categorizedunder two broad aspects of death: themes around the event ofdeath and themes about the pain and suffering of death. Underthe first broad category were fears regarding the uncertainty ofdeath (how, when), the unknown afterlife, the finality that the event

FIGURE 3 Interaction between fear of own dying and other CLFD scales.

Fear of Death and Loss of Loved Ones 413

brings, and the possibility of an unfulfilled life. In the second broadcategory of pain and suffering were fears of physical pain and ofthe emotional pain of leaving others behind in death or the deathof loved ones. Each theme is discussed in turn, followed by a dis-cussion of the differences in responses between the low and highfear of own death groups.

UNCERTAINTY OF DEATH

In terms of the event of death itself, some respondents werefearful of the uncertainty surrounding one’s death. Part of this fearinvolved not knowing how one will die, particularly as it mightinvolve a traumatic event like a heart attack or accident. Anotherbasis of this fear was uncertainty about when we die. Thus, theunpredictability of death, and therefore life, seemed to underpinmost respondents’ fears concerning this theme (e.g., ‘‘it could hap-pen to anyone, anywhere, without warning’’ and ‘‘being there oneminute, and gone the next’’).

Conversely, a minority of respondents feared the possibility ofhaving a predictable, certain death. That is, what they found mostfearsome was the possibility that they may know in advance thatthey were dying. For example, being a cancer sufferer and beinggiven a prognosis of a certain amount of time to live was seen asmore fearsome than a sudden unforseen death: ‘‘The thought ofhaving something and knowing you have this long to live scaresme.’’

UNKNOWN AFTERLIFE

Similar to the fearsome uncertainties surrounding the event ofdeath itself, many respondents also expressed fears regarding notbeing certain about what happens after death. Aspects of this themeincluded not knowing whether heaven and=or hell really exist, andwhether we continue to exist in some way after death either phy-sically or spiritually. For example, ‘‘The unknown . . . that theremay not be anything else for us’’; ‘‘It’s just the unknown, do wego somewhere or do we disappear’’; and ‘‘Where does your soulgo? Is that it? What happens to my body once I’m dead?’’

FINALITY OF DEATH

So while some respondents were fearful of an unknown after-life, others had no doubt about the afterlife but were fearful of the

414 D. M. Bath

perceived nothingness that death brings, the absolute finality ofdeath. For some respondents it was difficult to accept that, withdeath, ‘‘life’s just over.’’ For others, it was almost incredible to con-sider the end to their physical and spiritual existence, their mindand body, as well as to their existence in society. For these indivi-duals, the futility and pointlessness of life was part of what theyfeared. For example, ‘‘how suddenly you don’t think or feel any-more’’; ‘‘no longer thinking, breathing, and being forgotten’’;and ‘‘we love so deeply and struggle so much but in the end it’sall gone.’’

UNFULFILLED LIFE

Related to the idea that death is completely final, respondentsalso expressed fears that life may be unfulfilled at the time of death,particularly if death comes at a relatively young age. The idea ofdying before having accomplished what is important to the indi-vidual was central to this theme. For example, ‘‘that I will diebefore I am ready and have done everything I want to do in life,’’‘‘thoughts of regret of a wasted life,’’ and ‘‘to die and have notaccomplished or experienced something I find of value.’’

PAIN OF DEATH

Whilst a few respondents did not fear their own death, manyothers fear the possibility of a physically painful death. This painwas most feared if it were to be extreme in intensity or of long dur-ation. For example, ‘‘the thought of a painful slow death,’’ ‘‘that Iwill . . . experience intense amounts of pain before I actually die,’’and ‘‘dying in pain.’’

LOSS OF LOVED ONES

For many respondents, what they feared about death relatedto the pain of loss or detachment from loved ones: either in antici-pation of leaving loved ones behind when they die (‘‘not seeingeveryone I love anymore’’ and ‘‘not seeing the important thingsin the lives of my family’’); or, on the other hand, the pain of losswhen a loved one dies (‘‘the loss of a loved to me is almost unbear-able’’ and ‘‘my own death doesn’t worry me as much as the deathof someone that is close to me’’); or finally, the fear that their owndeath would create pain and suffering for loved ones (‘‘the pain

Fear of Death and Loss of Loved Ones 415

and suffering of the ones left behind’’ and ‘‘I am afraid to causepain to others if I die’’).

High vs. Low Fear of Death Groups

In order to explore possible differences in the relative emphasesplaced on themes between those with high and low fear of death,participants were grouped according to mean scores for the com-bined Death and Dying of Self scales, and each group was also splitaccording to gender. Those scoring above the mean were groupedas high fear and those scoring below the mean were grouped as lowfear. Chi-square analyses were conducted separately for men andwomen by each theme, where fear of death (high vs. low) wasthe grouping variable. This process was also conducted based onthe mean scores for the combined Death and Dying of Othersscales. Frequencies of responses are presented in Table 2. Therewere equal numbers of women in high and low fear groups forboth analyses. However for men, there were equal numbers ofparticipants across high and low Death and Dying of Self groups,but fewer men in the low fear group than in the high fear groupfor Death and Dying of Others (30% and 70%, respectively). Onlytwo analyses revealed significant group differences (see Table 2),however, in some cases this was most likely due to a lack of powerfrom small and unequal group sizes (i.e., only 5 men in the low fearof others group).

In terms of uncertainty of death, the frequency of responseswas similar across Death=Dying of Self and Other groups, butmore prevalent for high fear than low fear participants. Overall,this was the least common theme. Finality of death was the secondleast common theme, and overall, the frequency of responses wassimilar across both self and others analyses. No men, and fewwomen, in both low fear groups reported this theme. Howeverin the high fear groups, there were similar frequencies of responsesto the unknown afterlife and unfulfilled life categories (approxi-mately 25% of participants).

The unknown afterlife was a moderately common theme andagain, overall frequencies of responses were similar across both selfand others analyses. However, women with high fear of owndeath=dying were less likely than those with high fear of othersdeath=dying to respond with this theme. An unfulfilled life was

416 D. M. Bath

TABLE

2Frequ

ency

ofQua

litativeRespon

sesRegardingFears

ofDeath

byHighan

dLow

Fearof

Death

andGen

der

%of

participan

tswithrespon

ses

Group

Uncertainty

ofdeath

Unkn

own

afterlife

Finality

ofdeath

Unfulfilled

life

Pain

ofdeath

Lossof

lovedon

es

Owndeath

anddying

––

––

––

Highfear:Men

(n¼8)

12.50

25.00

20.0

25.0

20.0

25.0

Highfear:W

omen

(n¼32

)21

.90

15.60

28.1

a21

.918

.837

.5Low

fear:Men

(n¼9)

0.00

22.20

0.0

33.3

33.3

55.6

Low

fear:W

omen

(n¼31

)9.70

32.30

6.5�

19.4

32.3

57.6

Total

13.75

23.75

17.5

22.5

26.3

43.8

Other

death

anddying

Highfear:Men

(n¼12

)8.30

25.00

25.0

25.0

25.0

33.3

Highfear:W

omen

(n¼32

)21

.90

25.00

21.9

21.9

15.6

b37

.5Low

fear:Men

(n¼5)

0.00

20.00

0.0

40.0

40.0

60.0

Low

fear:W

omen

(n¼31

)7.60

22.20

12.9

19.4

35.5

b51

.6Total

13.80

23.75

17.5

22.5

26.3

43.8

Note.Chi-squa

rean

alyses

withsign

ifican

tgrou

pdifferences;

a v2(1)¼

5.31

,p¼.023

,b v

2(1)¼

3.28

,p¼.056

.

417

also a moderately common theme overall. Generally, there weresimilar frequencies of responses across groups, except for menwho were low in fear of death=dying, who were the most likelyto report this theme. Similarly, the pain of death was a moderatelycommon theme. Generally, similar frequencies of responses wereevident across high fear of death=dying groups. However, bothmen and women in the low fear of death=dying groups were morelikely to report this theme.

Consistent with the quantitative results discussed above, show-ing that individuals tend to fear others’ death or dying regardless ofthe degree to which they fear their own death, the loss of lovedones theme was the most prevalent amongst all respondents withalmost 50% of respondents making a comment in this category.Except for the high fear of own death=dying in men, this categorywas the most common for all other groups and generally it wasmore likely to occur in the low fear of death=dying groups. Thisis most likely due to the fact that, whereas high fear of death part-icipants tended to report a variety of fears across several themes(and thus responses were more dispersed), often the loss of lovedones was the only theme commented on by participants in thelow fear groups. Indeed, all of the respondents in the low feargroups who stated that they did not fear their own death didexpress fears regarding the effect their death would have on others,and=or the pain they would experience with the loss of a lovedone. For some, this stance involved having a very clear view onthe natural course of life, of which death is simply a part, whereasfor others this wasn’t necessarily bound in a positive view of oracceptance of death, but rather was based on the contemplationof something far worse, the death of a loved one.

Discussion

The present study explored the relationship between the fear ofone’s own death and dying and that of others, using participantresponses on the CLFD and to the question, ‘‘Which aspects ofdeath do you find the most worrying or fearsome?’’ Before discuss-ing the findings in relation to this aim, it is first important to notewhether the responses in the present study are similar to other gen-eral participant groups used in previous research, as it is essential

418 D. M. Bath

that any additional findings are not simply an artifact of thisparticular group.

The mean scores on the CLFD scales are consistent with priorresearch, in terms of both the magnitude of scores and the patternamongst the four scales, with the Death of Others and Dying ofOthers scores being the highest and the Death of Self and Dyingof Self scores being the lowest (e.g., Mooney & O’Gorman,2001). Scale reliabilities were also consistent with previous reports,showing a moderately high degree of internal consistency(Abdel-Khalek & Lester, 2004; Lester, 1994). Gender differenceswere found, and these were also consistent with other studies, withwomen showing a higher level of fear of death in general com-pared to men (Lester, 1994; Loo & Shea, 1996).

The fear of death themes identified in the qualitative data arealso consistent with previous research and the items or dimensionsincluded in measures developed from that research; for example,the Multidimensional Orientation Toward Dying and DeathInventory (MODDI-F; Wittkowski, 2001), the Fear of PersonalDeath Scale (FPDS; Florian & Karvetz, 1983), the Leming Fearof Death Scale (LFDS; Leming & Dickinson, 1985), and the Rea-sons for Death Fear Scale (RDFS; Abdel-Khalek, 2002b). Specifi-cally, the FPDS contains many of the themes that emerged fromthe present study, including the loss of self-fulfilment, self-annihilation (e.g., the cessation of the ability to think and experi-ence), transcendental consequences (e.g., uncertainty of afterlife),and the severing of ties with loved ones. Three of the four RDFSfactors also resonate with the current themes, including fear of painand punishment, fear of losing worldly involvements, and partingfrom loved ones. Similarly, the majority of scales included in theLFDS also map to the key themes identified in the present study:fear of pain, fear of isolation, separation and loneliness, fear ofafterlife concerns, fear of the finality of death, and fear of leavingloved ones.

The consistency of data reported here with previous researchprovides an indication that the current sample of respondents donot represent an unusual group, and therefore the findings maybe interpreted with some confidence. As stated previously, therelationships amongst the CLFD scales representing fear of one’sown death and dying and that of others have typically beenexamined using correlational analyses. These indicate that the

Fear of Death and Loss of Loved Ones 419

relationships are positive and linear such that, for example, as fearof own death increases, so does fear of others’ death. The results ofthe present study however have demonstrated that although thiskind of relationship is very clear for the death and dying scaleswithin the self dimension, and also within the others dimension,it is not so clear between these two dimensions themselves.

Whilst correlations between, for example, Death of Self andDeath of Others scales and Death of Self and Dying of Othersscales indicate moderate to strong positive relationships, chi-squareanalyses show a more elaborate set of patterns. As would beexpected given correlations between the scales, those who havea high fear of their own death will also most likely have a high fearof others’ death, and a high fear of others’ dying. However, unlikethe correlations indicate, those who have a low fear of their owndeath are almost equally likely to be high or low in fear of othersdeath or others dying (rather than being more likely to be low infear of others death or dying). So if magnitude of fear is not theonly element that explains the relationship between fear of owndeath=dying and fear of others death=dying, what other factorsare important?

In asking participants to respond to the open-ended questionabove, it was hoped to elicit the things about death that first cometo mind as being fearsome, and that these data would help toexplain the relationship between fear of own death and fear ofothers’ death. As research has already shown that people’s fearof death is multidimensional in nature it was not surprising thatresponses to this question were varied, and indeed, as alreadynoted, the themes that did emerge from the data were similar tothose already discussed in the literature. However, considerationof these themes in relation to high and low fear of own death, aswell as gender, has provided a basis for further elaboration of thisrelationship. First, having little fear of one’s own death does notrequire having little fear of death in general; all respondents in thiscategory indicated that they did fear some aspects of death but thatthese were related to the loss of loved ones or the pain that lovedones might suffer as a result of one’s own death. This makes sensein light of the chi-square analyses indicating that many with a lowfear of their own death actually have a high fear of others deathand=or dying. Loss of loved ones, through one’s own death orthe death of others, was not only the most prevalent theme for

420 D. M. Bath

the low fear group, but it was also the most frequent response forthe high fear group. Again, given the quantitative data, this is notsurprising. Mean scores on the Death of Others and Dying ofOthers scales were higher than the Death of Self and Dying of Selfscales, and in practical terms, the chi-square analyses indicate thatit is likely that more people have a high fear of others’ death anddying than a high fear of their own death. Moreover, both womenand men scored highest on the Death of Others scale and the dif-ference between genders was least marked for this scale comparedto the other CLFD scales.

Gender differences in death anxiety research have tended tobe inconclusive and sometimes contradictory, where, for example,some studies show that overall women fear death more than menwhereas others report no differences (e.g., Neimeyer, 1994). Thefindings here are no different in this respect. For example, incon-sistent with the present study, previous research has reported thatwomen show greater fear of pain in relation to death than do men(Cicirelli, 1998). Also, in a study of personal meanings of death,Holcomb, Neimeyer, and Moore (1993) found that women weremore likely than men to view death as involving high certaintyand continued existence, whereas in the present study womenwere more likely than men to worry about the possible uncertaintyof death and the finality of death. Further, in his study of personalmeanings of death and fear of death, Cicirelli (1998) found thatwomen were more likely than men to view death in terms of thefinality of existence. In addition, he found no gender differencesregarding the fear of leaving loved ones. Although the currentstudy is consistent with this latter finding, it is inconsistent withother research which has found that men tend to view death ashaving less personal impact on them, in comparison to womenwho view death in highly emotional and negative ways (e.g.,Holcomb et al., 1993). It is likely, as suggested by others, thatrather than gender itself being an influential factor on death anxi-ety, such gender differences may in fact be driven by other relatedvariables, such as locus of control (Neimeyer, 1994) or attachmentstyle (Beggs, 2000; Waskowic & Chartier, 2003).

There are a number of issues with the present study thatshould be considered, however, in the interpretation of theseresults. First, as is typical of much research of this nature, thesample of respondents does not include the full spectrum of age

Fear of Death and Loss of Loved Ones 421

groups, nor does it have equal numbers of men and women.Obviously, as age and gender have previously been found to influ-ence death anxiety by some studies (Neimeyer et al., 2004), it isimportant that future work attempts to investigate the relationshipbetween fear of own death and fear of others’ death with a broaderage range and more equal proportions of men and women. Also,the qualitative data presented here were derived from questionnaireresponses to an open-ended question rather than from in-depthinterviews, and as such a richer and more thorough explorationof the meanings underpinning individuals’ responses could not beexamined. In order to fully understand individual’s fears aboutdeath, a more in-depth approach needs to be taken. Also, there isalways the possibility that participants’ responses were influencedby the ordering of the survey components. Future research shouldconsider a counterbalanced design or alternative ordering of ques-tions in order to further establish the validity of the present findings.

Together, however, the results of the present study pointtowards one key finding—individuals tend to fear the death ofothers regardless of the degree to which they fear their own death,and that the detachment or separation from others through death(either one’s own death or the death of another) is one of the mostfearsome aspects of death, a fear that is shared by the majority ofrespondents. It is no wonder, therefore, that continuing the bondbetween living and deceased has become an important theme inthe bereavement literature and it is now recognized that maintain-ing some degree of attachment is a normal part of adjusting to theloss of a loved one (Field, 2006a). Various cultures and religionshave evolved beliefs and customs relating to the ongoing relation-ship with the dead for many centuries, but current media trendsalso appear to indicate increasing interest in the general publicfor continuing to have contact with loved ones who are deceased.For example, a plethora of television dramas and talk-show pro-grams feature mediums claiming to connect with the other side(e.g., Edward, 2003).

Perhaps the findings of the present study demonstrate oneway in which research on death anxiety can inform research andpractice in grief and bereavement. For instance, although continu-ing bonds in bereavement is acknowledged to be normative andadaptive, recent research has also discussed conditions where thismay in fact be maladaptive (e.g., Field, 2006b; Klass, 2006).

422 D. M. Bath

Neimeyer, Baldwin, and Gillies (2006) suggested that the greatestdifficulties in grieving arise when the individual has an inabilityto make sense of the loss of their loved one. The ability to makesense of the loss may be greatly influenced by an individual’sbeliefs and fears about death, and so perhaps further knowledgeabout the fears associated with death, in particular the leaving, orloss, of loved ones as indicated by the present study, may assistfuture developments in therapeutic interventions. Indeed, Field(2006a) suggested that it is necessary for research to continue toidentify and elucidate the kinds of interventions needed to pro-mote healthy continuing bonds in the bereaved.

References

Abdel-Khalek, A. M. (2002a). Convergent validity of the Collet–Lester Fear ofDeath and Templer Death Anxiety Scales in Egyptian male college students.Psychological Reports, 90, 849–850.

Abdel-Khalek, A. M. (2002b). Why do we fear death? The construction andvalidation of the Reasons for Death Fear Scale. Death Studies, 26, 669–680.

Abdel-Khalek, A. M., & Lester, D. (2004). The factorial structure of the Arabicversion of the Revised Collett–Lester Fear of Death Scale. Death Studies,28, 787–794.

Beggs, J. R. (2000). The relationship between attachment style and death anxiety asrelated to a romantic partner. Unpublished doctoral dissertation, Georgia StateUniversity, Atlanta, Georgia.

Cicirelli, V. G. (1998). Personal meanings of death in relation to fear of death.Death Studies, 22, 713–833.

Claxton-Oldfield, S., Crain, M., & Claxton-Oldfield, J. (2007). Death anxiety anddeath competency: The impact of a palliative care volunteer trainingprogram. American Journal of Hospice and Palliative Medicine, 23, 464–478.

Clements, R. (1998). Intrinsic religious motivation and attitudes towards deathamong the elderly. Current Psychology: Developmental, Learning, Personality,Social, 17, 237–248.

Collett, L., & Lester, D. (1969). The fear of death and dying. Journal of Psychology,72, 179–181.

Edward, J. (2003). After life: Answers from the other side. New York: Princess Books.Field, N. P. (2006a). Continuing bonds in adaptation to bereavement:

Introduction. Death Studies, 30, 709–714.Field, N. P. (2006b). Unresolved loss and maladaptive continuing bonds expres-

sions. Death Studies, 30, 739–756.Florian, V., & Kravetz, S. (1983). Fear of personal death: Attribution, structure

and relation to religious belief. Journal of Personality and Social Psychology, 44,600–607.

Fear of Death and Loss of Loved Ones 423

Handal, P. J. (1975). Relationship between the Death Anxiety Scale andrepression. Journal of Clinical Psychology, 31, 675–677.

Hayes, N. (2002). Introducing quantitative analysis. In N. Hayes (Ed.), Doingpsychological research: Gathering and analyzing data (pp. 168–182).Buckingham, UK: Open University Press.

Hayslip, B., & Walling, M. L. (1985–1986). Impact of hospice volunteer trainingon death anxiety and locus of control. Omega: The Journal of Death and Dying,16, 243–254.

Hoelter, J. W. (1979). Multidimensional treatment of fear of death. Journal ofConsulting and Clinical Psychology, 47, 996–999.

Holcomb, L. E., Neimeyer, R. A., & Moore, M. K. (1993). Personal meanings ofdeath: A content analysis of free-response narratives.Death Studies, 17, 299–318.

Jordan, T. J., Ellis, R. R., & Grallo, R. (1986). A comparison of levels of anxiety ofmedical students and graduate counsellors about death. Journal of MedicalEducation, 61, 923–925.

Klass, D. (2006). Continuing conversations about continuing bonds. Death Studies,30, 843–858.

Knight, K., Elfenbein, M., & Capozzi, L. (2000). Relationship of recollections offirst death experience to current death attitudes. Death Studies, 24, 201–221.

Larrabee, M. (1978). Measuring fear of death: A reliability study. Journal ofPsychology: Interdisciplinary and Applied, 100, 33–37.

Leming, M. R., & Dickinson, G. E. (1985). Understanding dying, death and bereave-ment. New York: Holt, Rinehart & Winston.

Lester, D. (1972). Studies in death attitudes. Psychological Reports, 30, 440.Lester, D. (1984–1985). The fear of death, sex and androgyny. Omega: The Journal

of Death and Dying, 15, 271–274.Lester, D. (1994). The Collett–Lester Fear of Death Scale. In R. A. Neimeyer

(Ed.), Death anxiety handbook: Research, instrumentation and application (pp.45–60). New York: Taylor & Francis.

Lester, D. (2004). The factorial structure of the revised Collett–Lester Fear ofDeath Scale. Death Studies, 28, 795–779.

Lester, D., & Abdel-Khalek, A. M. (2003). The Collett–Lester Fear of Death Scale:A correction. Death Studies, 27, 81–85.

Loo, R. (1984). Personality correlates of the fear of death. Journal of ClinicalPsychology, 40, 120–122.

Loo, R., & Shea, L. (1996). Structure of the Collett–Lester Fear of Death andDying Scale. Death Studies, 20, 577–586.

McMordie, W. R. (1979). Improving measurement of death anxiety. PsychologicalReports, 44, 975–980.

Mooney, D. C. (2005). Tactical reframing to reduce death anxiety in undergrad-uate nursing students. American Journal of Hospice and Palliative Medicine, 22,427–432.

Mooney, D. C., & O’Gorman, J. (2001). Construct validity of the revised Collett–Lester Fear of Death and Dying Scale. Omega: The Journal of Death and Dying,43, 157–173.

Neimeyer, R. A. (1994). Death anxiety handbook: Research, instrumentation andapplication. New York: Taylor & Francis.

424 D. M. Bath

Neimeyer, R. A., Bagley, K. J., & Moore, M. K. (1986). Cognitive structure anddeath anxiety. Death Studies, 10, 273–728.

Neimeyer, R. A., Baldwin, S. A., & Gillies, J. (2006). Continuing bonds and recon-structing meaning: Mitigating complications in bereavement. Death Studies,30, 715–738.

Neimeyer, R. A., & Dingemans, P. M. (1980–1981). Death orientation in the sui-cide intervention worker. Omega: The Journal of Death and Dying, 11, 15–23.

Neimeyer, R. A., Moser, R., & Wittkowski, J. (2003). Assessing attitudes towarddying and death: Psychometric considerations. Omega: The Journal of Deathand Dying, 47, 45–76.

Neimeyer, R. A., Wittkowski, J., & Moser, R. (2004). Psychological research ondeath attitudes: An overview and evaluation. Death Studies, 28, 309–340.

Robbins, R. (1990). Signing an organ donor card: Psychological factors. DeathStudies, 14, 219–229.

Robbins, R. (1990–1991). Bugen’s Coping with Death Scale: Reliability andfurther validation. Omega: The Journal of Death and Dying, 22, 287–299.

Smith, D. K., Nehemkis, A. M., & Charter, R. A. (1983–1984). Fear of death,death attitudes and religious conviction in the terminally ill. InternationalJournal of Psychiatry in Medicine, 13, 221–232.

Soghomonian, G. (2000). The influence of fear of death and psychologists’attitudes toward working with the elderly. Unpublished doctoral dissertation,California School of Professional Psychology, Alameda, California.

Straub, S., & Roberts, J. M. (2001). Fear of death in widows: Effects of age atwidowhood and suddenness of death. Omega: The Journal of Death and Dying,43, 25–41.

Templer, D. I. (1970). The construction and validation of a death anxiety scale.Journal of General Psychology, 82, 165–177.

Tomas-Sabado, J., Limonero, J. T., & Abdel-Khalek, A. M. (2007). Spanish adap-tation of the Collett–Lester Fear of Death Scale. Death Studies, 31, 249–260.

Vargo, M. (1980). Relationship between the Templer Death Anxiety Scale andthe Collett–Lester Fear of Death Scale. Psychological Reports, 46, 561–562.

Vickio, C. J., Cavanaugh, J. C., & Attig, T. W. (1990). Perceptions of grief amonguniversity students. Death Studies, 14, 231–240.

Waskowic, T. D., & Chartier, B. M. (2003). Attachment and the experience ofgrief following the loss of a spouse. Omega: The Journal of Death and Dying,47, 77–91.

Wittkowski, J. (2001). The construction of the Multidimensional OrientationToward Dying and Death Inventory (MODDI-F). Death Studies, 25, 479–495.

Woodford, L. M. (2002). Sharing the pain with friends: Grief and support among lateadolescent women. Unpublished doctoral dissertation, Queen’s UniversityKingston, Ontario, Canada.

Yasukawa, N. (2006). The factorial structure of the Japanese version of the revisedCollett–Lester Fear of Death Scale. Unpublished masters dissertation, GriffithUniversity, Brisbane, Australia.

Fear of Death and Loss of Loved Ones 425

Copyright of Death Studies is the property of Routledge and its content may not be copied or emailed to

multiple sites or posted to a listserv without the copyright holder's express written permission. However, users

may print, download, or email articles for individual use.